The identification of aerodynamic parameters is accomplished through the test data of the dynamic movement of scaled aircraft models flying dynamically in wind tunnel, which can realize the accurate acquisition of the aerodynamic model of the aircraft in the preliminary stage for aircraft design, and it is of great significance for improving the efficiency of aircraft design. However, the translational motion of the test model in the wind tunnel virtual flight is subject to constraints that result in distinct flight dynamics compared to free flight. These constraints have implications for the accuracy of aerodynamic derivatives obtained through the identification of wind tunnel test data. With this issue in mind, the research studies the differences in longitudinal dynamic characteristics between unconstrained free flight and wind tunnel virtual flight, and innovatively proposes an online correction test based wind tunnel virtual flight test technique. The longitudinal trajectory and velocity changes of the model are solved online by the aerodynamic forces measured during the test, and then the coupled relationship between aircraft translation and rotation is used to correct the model’s pitch attitude motion online. For the first time, the problem of solving the data approximation for free flight has been solved, eliminating the difference between the dynamics of wind tunnel virtual flight and free flight, and improving the accuracy of the aerodynamic derivative identification results. The experiment’s findings show that accurate aerodynamic derivatives can be identified based on the online correction test data, and the observed behaviour of the identified motion model has similarities to that of the free flight motion model.